{"id":41638,"date":"2025-08-23T11:49:02","date_gmt":"2025-08-23T03:49:02","guid":{"rendered":"https:\/\/www.1ai.net\/?p=41638"},"modified":"2025-08-23T11:49:02","modified_gmt":"2025-08-23T03:49:02","slug":"%e9%92%89%e9%92%89%e8%81%94%e5%90%88%e9%80%9a%e4%b9%89%e6%8e%a8%e5%87%ba-fun-asr-%e8%af%ad%e9%9f%b3%e8%af%86%e5%88%ab%e5%a4%a7%e6%a8%a1%e5%9e%8b%ef%bc%8c%e8%83%bd%e5%90%ac%e6%87%82%e5%ae%b6%e8%a3%85","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/41638.html","title":{"rendered":"Nail joint Tongyi launched Fun-ASR speech recognition big model, can understand home decoration, animal husbandry and other top ten industry black words"},"content":{"rendered":"

August 23rd.DingTalk<\/a>, the Tongyi Labs voice team announced yesterday that it has joined forces to launch theNew GenerationSpeech recognition macromodel<\/a> Fun-ASR<\/a><\/strong>I can hear you.Top 10 industry hacks for home improvement, livestock and more<\/strong>and supports enterprise-specific model customization training. Currently, Fun-ASR has been integrated into several functional modules of Nail, including meeting captioning and simultaneous transmission, intelligent minutes, voice assistant and so on.<\/p>\n

\"409d77c8j00t1fh89005hd000u000gwp\"<\/p>\n

1AI has learned that Fun-ASR is now integrated into several functional modules of Nail, includingMeeting Captioning & Simultaneous Interpreting, Intelligent Minutes, Voice Assistants<\/strong>etc., aiming to build a set of stable, efficient, and easy-to-scalable speech recognition bases, especially suitable for scenarios with high requirements for enterprise-level context understanding and recognition accuracy.<\/p>\n

According to the official description, the model has undergoneHundreds of millions of hours<\/strong>Audio data trained to accurately understandInternet, technology, home improvement, livestock, automotive<\/strong>and more than a dozen other areas of specialized terminology.<\/p>\n

The actual test data shows that Fun-ASR's accuracy in the insurance industry has improved by 18% compared with the previous one, and it has also achieved 15%-20% in the home decoration and animal husbandry industries. For out-of-the-way words, Fun-ASR alsoProvide the ability for companies to customize hot words<\/strong>It supports importing up to 1000+ hot words.<\/p>\n

Fun-ASR can combine the enterprise's pre-existing information within Nail such as address book, calendar, knowledge base, etc., to perform inference optimization, effectively mitigating the illusion caused by large models and providing more reliable transcription results. This capability requires authorization from the enterprise before it can take effect.<\/p>\n

Nail provides a use scenario: Take the cooperation between Gujia Home Furnishing and Nail as an example, after exclusive training, when Gujia's store sales explain the products to customers, the model can accurately identify \"Pulse Pulse Latex Imported from Belgium\", \"Original Sonocore Foam Process\" and so on. \"etc.Specialized expressions<\/strong>The following is a reliable basis for further analysis of customer needs.<\/p>","protected":false},"excerpt":{"rendered":"

August 23 news, nail, Tongyi Labs voice team announced yesterday, joined forces to launch a new generation of speech recognition big model Fun-ASR, can understand home improvement, animal husbandry and other ten industry black words, and support enterprise exclusive model custom training. At present, Fun-ASR has been integrated into a number of functional modules of Nail, including meeting subtitles and simultaneous transmission, intelligent minutes, voice assistant and so on. 1AI learned that Fun-ASR has been integrated into several functional modules of Nail, including meeting subtitling and simultaneous transmission, intelligent minutes, voice assistant, etc. It aims to build a set of stable, efficient, and easily scalable speech recognition base, which is especially suitable for the scenarios that have high requirements for enterprise-level context understanding and recognition accuracy. According to the official introduction, the model has undergone hundreds of millions of hours of audio data training, and is able to accurately<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[7459,4828,349],"collection":[],"class_list":["post-41638","post","type-post","status-publish","format-standard","hentry","category-news","tag-fun-asr","tag-4828","tag-349"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/41638","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=41638"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/41638\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=41638"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=41638"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=41638"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=41638"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}